| title | Surflo: Consistent 3D Surface Flow Model with Global State |
|---|---|
| authors | Antoine Guédon, Shu Nakamura, Nicolas Dufour, Jiahui Lei, Ko Nishino, Angjoo Kanazawa |
| date | 2026-06-11 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2606.13644 |
| upvotes | 8 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | Geometry is invariant to viewpoint, which makes any collection of images a redundant encoding of a single 3D state. Existing feed-forward reconstruction models fail to exploit this: per-view methods emit overlapping, unaligned pointmaps that grow linearly with input count, while global-latent methods commit to a fixed, low-resolution output. We introduce Surflo, which compresses a variable number of unposed RGB views into K latent tokens-one global state-and decodes oriented 3D surface points by independently transporting them from noise onto the surface via flow matching. This frees the output from any fixed grid or token budget: the same latent yields from a few thousand to a million points in a single forward pass. To suppress the local inconsistencies inherent to independent per-point decoding, an inference-time guidance term correlates nearby points by injecting a photometric gradient during ODE integration. Surflo matches or surpasses feed-forward baselines on surface metrics, runs an order of magnitude faster than optimization-based methods that require hundreds of views, and is the only feed-forward approach to combine a global latent with arbitrary-resolution decoding. |
cite
@misc{gudon2026surflo,
title={Surflo: Consistent 3D Surface Flow Model with Global State},
author={Antoine Guédon and Shu Nakamura and Nicolas Dufour and Jiahui Lei and Ko Nishino and Angjoo Kanazawa},
year={2026},
eprint={2606.13644},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2606.13644}
}